An impaired healing model of osteochondral defect in papain-induced arthritis

Abstract

Background: Osteochondral defects (OCD) are common in osteoarthritis (OA) and difficult to heal. Numerous tissue engineering approaches and novel biomaterials are developed to solve this challenging condition. Although most of the novel methods can successfully treat osteochondral defects in preclinical trials, their clinical application in OA patients is not satisfactory, due to a high spontaneous recovery rate of many preclinical animal models by ignoring the inflammatory environment. In this study, we developed a sustained osteochondral defect model in osteoarthritic rabbits and compared the cartilage and subchondral bone regeneration in normal and arthritic environments.

Methods: Rabbits were injected with papain (1.25%) in the right knee joints (OA group), and saline in the left knee joints (Non-OA group) at day 1 and day 3. One week later a cylindrical osteochondral defect of 3.2 mm in diameter and 3 mm depth was made in the femoral patellar groove. After 16 weeks, newly regenerated cartilage and bone inside the defect were evaluated by micro-CT, histomorphology and immunohistochemistry.

Results: One week after papain injection, extracellular matrix in the OA group demonstrated dramatically less safranin O staining intensity than in the non-OA group. Until 13 weeks of post-surgery, knee width remained significantly higher in the OA group than the non-OA control group. Sixteen weeks after surgery, the OA group had 11.3% lower International Cartilage Regeneration and Joint Preservation Society score and 32.5% lower O'Driscoll score than the non-OA group. There were less sulfated glycosaminoglycan and type II collagen but 74.1% more MMP-3 protein in the regenerated cartilage of the OA group compared with the non-OA group. As to the regenerated bone, bone volume fraction, trabecular thickness and trabecular number were all about 28% lower, while the bone mineral density was 26.7% higher in the OA group compared to the non-OA group. Dynamic histomorphometry parameters including percent labeled perimeter, mineral apposition rate and bone formation rate were lower in the OA group than in the non-OA group. Immunohistochemistry data showed that the OA group had 15.9% less type I collagen than the non-OA group.

Conclusion: The present study successfully established a non-self-healing osteochondral defect rabbit model in papain-induced OA, which was well simulating the clinical feature and pathology. In addition, we confirmed that both cartilage and subchondral bone regeneration were further impaired in arthritic environment.

The translational potential of this article: The present study provides an osteochondral defect in a small osteoarthritic model. This non-self-healing model and the evaluation protocol could be used to evaluate the efficacy and study the mechanism of newly developed biomaterials or tissue engineering methods preclinically; as methods tested in reliable preclinical models are expected to achieve improved success rate when tested clinically for treatment of OCD in OA patients.

Keywords: Impaired healing; Osteochondral defect; Papain-induced osteoarthritis.

Fig. 1

Papain-induced OA caused severe cartilage damage in the pilot study. A: Diagram for establishing an OCD model in papain-induced OA rabbits in the pilot study. B: Macroscopic observation revealed that the cartilage surface in the OA group was severely damaged by papain 12 weeks after surgery, white arrow: rough surface indicating damaged cartilage. C: Histological analysis (S–F: safranin O-fast green, T–B: toluidine blue and H&E: hematoxylin and eosin) revealed that the cartilage self-repair was limited in both papain-induced OA and non-OA rabbits, black arrows: the margins of the defect. D: Micro-CT analysis showed less subchondral bone formation in the defect region. E: the change of knee width over time showed chronic swollen knee existed in the OA group. Data are shown as mean ​± ​SEM, N ​= ​5. Two-way ANOVA followed by Holm-Sidak’s multiple comparison test were used to evaluate differences between OA and non-OA groups at the same time point. ∗p ​< ​0.05.

Fig. 2

Cartilage damage and chronic swollen knee in OA rabbits. A: Diagram for establishing an OCD model in papain-induced OA rabbits. On the tenth and seventh days before surgical operation, papain or saline was articularly injected into the right or left knee, respectively. Then, OCD (3.2 ​mm diameter ​× ​3.0 ​mm depth) was made in the femoral trochlear center. Knee width was recorded every week. Double fluorochrome labelling was performed by intra-muscular injection of calcein and xylenol orange disodium salt administered 10 and 3 days before sacrifice. 16 weeks after surgery, samples were harvested for analysis. B: Safranin O-fast green staining showed cartilage erosion and loss of proteoglycan in the OA group on the tenth day after the first injection of papain; C: macroscopic observation of the knee at 16 weeks after surgery, white arrow: rough surface indicating damaged cartilage; D: the change of knee width over time showed chronic swollen knee existed in the OA group. Data are shown as mean ​± ​SEM, N ​= ​8. Two-way ANOVA followed by Holm-Sidak’s multiple comparison test were used to evaluate differences between OA and non-OA groups at the same time point. ∗p ​< ​0.05.

Fig. 3

Histological analysis of cartilage repair at 16 weeks post-surgery. A: Representative histological images of newly formed tissue in OCD region at week 16 post-surgery. Safranin O-fast green (S–F), toluidine blue (T–B) and hematoxylin and eosin (H&E) staining in the non-OA group showed enhanced cartilage and subchondral bone repair, compared to the OA group. There is a large subchondral cyst in the non-OA group and OA group. The arrows denote the margins of the defect. B: International Cartilage Repair Regeneration and Joint Preservation Society (ICRS) scoring; C: The modified O’Driscoll histological scoring; D: The sGAG optical density. Data are shown as mean ​± ​SD, N ​= ​4, Paired t-test were used to evaluate differences between OA and non-OA groups. ∗p ​< ​0.05.

Fig. 4

Micro-CT analysis showed that the subchondral bone and cartilage formation in the defect area of the OA group was less than that of the non-OA group. A: The reconstructed images showed the joint surfaces, the newly formed bone in OCD region (new bone) and 2D image of the distal femur OCD site (red: cartilage). B–C: The quantitative micro-CT data of bone mineral density (BMD), bone volume fraction (BV/TV) of the host bone. D–H: The quantitative micro-CT data of bone mineral density (BMD), bone volume fraction (BV/TV), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp) and trabecular number (Tb.N) of newly formed bone. I: The X-ray attenuation value of the newly formed cartilage in OCD region. Data are shown as mean ​± ​SD, N ​= ​8. Paired t-test were used to evaluate differences between OA and non-OA groups. ∗p ​< ​0.05, ∗∗p ​< ​0.01.

Fig. 5

Dynamic histomorphometry of newly formed tissue in OCD region. A–C: Representative fluorescence images of subchondral bone in OCD region (A: Gross fluorescence image, B: Non-OA group, C: OA group). D–F: Quantification of percent labeling perimeter (%L.Pm), mineral apposition rate (MAR) and bone formation rate (BFR/BS). Data are shown as mean ​± ​SD, N ​= ​4. Paired t-test were used to evaluate differences between OA and non-OA groups. ∗p ​< ​0.05 and ∗∗p ​< ​0.01.

Fig. 6

Immunohistochemical (IHC) staining of collagen type I (COL I), collagen type II (COL II) and MMP-3 in newly formed tissue in OCD region. A: Representative images of IHC of COL I, COL II and MMP-3. B: The semi-quantitative results of COL I, COL II and MMP-3. Data are shown as mean ​± ​SD, N ​= ​4. Paired t-test were used to evaluate differences between OA and non-OA groups. ∗p ​< ​0.05 and ∗∗p ​< ​0.01.